DODD & FRANZ: UPLAND SNAKES


habitat preferences, may restrict the effectiveness of funnel trap sampling. Ford et
al. (1991) also were unable to capture Tantilla in funnel traps.
 Several species that were trapped, such as Nerodia fasciata and N.
 taxispilota, are aquatic species and as such were unexpected in xeric upland
 habitats. However, Dodd (1992) found some non-resident aquatic or wetland-
 associated species that regularly visited a small isolated temporary pond located in
 upland habitat on the Ordway Preserve. Snakes normally associated with wetland
 habitats may travel across unfavorable habitat to find foraging areas or to disperse
 during unfavorable environmental conditions (Dodd 1993; Seigel et al. ms). On the
 Ordway Preserve, six additional wetland-associated species (Farancia abacura,
 Nerodia floridana, Opheodrys aestivus, Regina alleni, Seminatrix pygaea,
 Thamnophis sauritus) are known to at least occasionally cross upland habitat
 (Dodd 1992; unpubl. observe ) but were not trapped during the study. The likelihood
 of trapping wetland-associated species as they move across upland habitat would
 seem to be small, unless the uplands were located adjacent to wetlands subject to
 periodic desiccation. In such locations, seasonal snake activity is influenced by
 fluctuations in the water table resulting in increased capture as wetlands dry
 (Bernardino and Dalrymple 1992).
 Finally, active foragers, such as Masticophis and Coluber, should be more
likely to encounter funnel traps than sit-and-wait predators such as Crotalus.
Active foragers, especially those that take a wide range of prey, also are more likely
than sit-and-wait predators to be drawn to traps through intra- or interspecific
chemical cues or the activity of prey species (lizards, other snakes, or rodents)
caught in the traps.
 Based on our results and those of other recent investigators (Fitch 1992;
Grant et al. 1992; Rodda and Fritts 1992), we suggest that funnel traps should not
be employed as the sole method for community sampling. All sampling techniques
have biases and limitations, but certain questions, such as those involving the
determination of activity patterns, often can be addressed using a non-trap biased
approach (Reynolds 1982; Price and LaPointe 1990; Dalrymple et al. 1991a;
Dalrymple et al. 1991b; Bernardino and Dalrymple 1992). Inventory sampling
should use a variety of techniques, such as pitfall traps with drift fences (Gibbons
and Semlitsch 1982), road-cruising (Klauber 1939), coverboards (Grant et al.
1992), and time-constraint sampling (Campbell and Christman 1982), to
supplement funnel trap data (Fitch 1992).


 LITERATURE CITED


Anderson Bell. 1987. ABSTAT. Release 4. Parker, Colorado.
Begon, M., J. L Harper, and C. R. Townsend. 1986. Ecology: Individuals, populations and communities.
 Blackwell, Oxford.
Bernardino, F. S., Jr., and G. H. Dalrymple. 1992. Seasonal activity and road mortality of the snakes of the
 Pa-hay-okee wetlands of Everglades National Park, USA. Biol. Conserve. 62:71-75.